Final analysis report

Bias, Coverage, MSE

Scenario 1.1

All treatments have equal effects: Tj = 37.5%, j = A, B, C, D

Scenario 1.2

Treatment B is better than treatments A, C and D: \(T_B = 35\%\); and \(T_j = 45\%\), \(j = A, C, D\)

Scenario 1.3

All 4 treatments have different effects (small difference): \(T_A = 30\%\); \(T_B = 35\%\); \(T_C = 40\%\); \(T_D = 45\%\)

Scenario 1.4

All 4 treatments have different effects (large difference): TA = 30%; TB = 40%; TC = 50%; TD = 60%

Scenario 2.1

Tj = 17.5%, j = A, B, C, D

Scenario 2.2

TA = 10%, TB = 15%, TC = 20%, TD = 25%

Scenario 2.3

Tj = 82.5%, j = A, B, C, D

Scenario 2.4

TA = 75%, TB = 80%, TC = 85%, TD = 90%

Scenario 3.1

Tj = 37.5%, j = A, B, C, D

Scenario 3.2

TA = 30%, TB = 35%, TC = 40%, TD = 45%

Scenario 4.1

Different effectiveness across patterns: Pattern 1: TA = 20%, TB = 25%, TC = 30%, TD = 35% Pattern 2: TA = 20% TB = 30% TC = 40% TD = 50% Pattern 3: TA = 40% TB = 60% TC = 75% TD = 95% Pattern 4: TA = 30% TB = 40% TC = 80% TD = 90%

Scenario 4.2

Different treatment effects in 8 sites compares to the other 2 sites: TAq = 30%, TBq = 35% TCq = 40% TDq = 45%, q = 3, …, 10; and TAq = 20%, TBq = 25%, TCq = 30%, TDq = 35%, q = 1, 2

Scenario 4.3

Different treatment effects in 6 sites compared to the other 4 sites, TAq = 30%. TBq = 35% TCq = 40% TDq = 45%, q = 5, …, 10; and TAq = 20%, TBq = 25%, TCq = 30%, TDq = 35%, q = 1, …, 4

Mortality gain

Scenario 1.2

Treatment B is better than treatments A, C and D: \(T_B = 35\%\); and \(T_j = 45\%\), \(j = A, C, D\)

Scenario 1.3

All 4 treatments have different effects (small difference): \(T_A = 30\%\); \(T_B = 35\%\); \(T_C = 40\%\); \(T_D = 45\%\)

Scenario 1.4

All 4 treatments have different effects (large difference): TA = 30%; TB = 40%; TC = 50%; TD = 60%

Scenario 2.2

TA = 10%, TB = 15%, TC = 20%, TD = 25%

Scenario 2.4

TA = 75%, TB = 80%, TC = 85%, TD = 90%

Scenario 3.2

TA = 30%, TB = 35%, TC = 40%, TD = 45%

Scenario 4.1

Different effectiveness across patterns: Pattern 1: TA = 20%, TB = 25%, TC = 30%, TD = 35% Pattern 2: TA = 20% TB = 30% TC = 40% TD = 50% Pattern 3: TA = 40% TB = 60% TC = 75% TD = 95% Pattern 4: TA = 30% TB = 40% TC = 80% TD = 90%

Scenario 4.2

Different treatment effects in 8 sites compares to the other 2 sites: TAq = 30%, TBq = 35% TCq = 40% TDq = 45%, q = 3, …, 10; and TAq = 20%, TBq = 25%, TCq = 30%, TDq = 35%, q = 1, 2

Scenario 4.3

Different treatment effects in 6 sites compared to the other 4 sites, TAq = 30%. TBq = 35% TCq = 40% TDq = 45%, q = 5, …, 10; and TAq = 20%, TBq = 25%, TCq = 30%, TDq = 35%, q = 1, …, 4

Power

Scenario 1.2

Treatment B is better than treatments A, C and D: \(T_B = 35\%\); and \(T_j = 45\%\), \(j = A, C, D\)

Scenario 1.3

All 4 treatments have different effects (small difference): \(T_A = 30\%\); \(T_B = 35\%\); \(T_C = 40\%\); \(T_D = 45\%\)

Scenario 1.4

All 4 treatments have different effects (large difference): TA = 30%; TB = 40%; TC = 50%; TD = 60%

Scenario 2.2

TA = 10%, TB = 15%, TC = 20%, TD = 25%

Scenario 2.4

TA = 75%, TB = 80%, TC = 85%, TD = 90%

Scenario 3.2

TA = 30%, TB = 35%, TC = 40%, TD = 45%

Scenario 4.1

Different effectiveness across patterns: Pattern 1: TA = 20%, TB = 25%, TC = 30%, TD = 35% Pattern 2: TA = 20% TB = 30% TC = 40% TD = 50% Pattern 3: TA = 40% TB = 60% TC = 75% TD = 95% Pattern 4: TA = 30% TB = 40% TC = 80% TD = 90%

Scenario 4.2

Different treatment effects in 8 sites compares to the other 2 sites: TAq = 30%, TBq = 35% TCq = 40% TDq = 45%, q = 3, …, 10; and TAq = 20%, TBq = 25%, TCq = 30%, TDq = 35%, q = 1, 2

Scenario 4.3

Different treatment effects in 6 sites compared to the other 4 sites, TAq = 30%. TBq = 35% TCq = 40% TDq = 45%, q = 5, …, 10; and TAq = 20%, TBq = 25%, TCq = 30%, TDq = 35%, q = 1, …, 4

Type 1 error

Scenario 1.1

Scenario 2.1

Scenario 2.3

Scenario 3.1

Running time